ARPA-H's July 10 SBIR Window Is a Different Animal — Seven Topics, a Contract Not a Grant, and a Solution Summary That Decides Everything

Most founders who have run the SBIR gauntlet at NIH or NSF arrive at ARPA-H expecting a familiar process and find something that does not behave the way they remember. The Advanced Research Projects Agency for Health was stood up to do for biomedicine what DARPA did for defense — fund the high-risk, high-reward swings that the traditional peer-review system is structurally unable to take. Its small-business program inherits that DNA, and the FY2026 SBIR/STTR solicitation is a clean illustration of the difference. The first gate — the Solution Summary — is due July 10, 2026 at 11:59 AM ET, and missing it means waiting for the next cycle.

This is not a deadline you cram for. ARPA-H runs a two-stage, gated process: a short Solution Summary first, and only if you clear that bar are you invited to deliver a full Technical Pitch and Cost Proposal, anticipated around September 9, 2026. That structure changes everything about how you should spend the time between now and the deadline. The mistake is to treat the summary as a formality on the way to the "real" application. At ARPA-H, the summary is the competition. Most of the field gets cut here.

Seven topics, and they are unusually specific

Unlike NSF's open-topic SBIR, where you propose whatever deep technology you can defend, ARPA-H publishes named topics that reflect concrete mission needs. The FY2026 solicitation (Notice ID 7599226SN106) names seven:

1. Annual fertility forecasting test — a diagnostic giving women predictive insight into future fertility, not just a current snapshot.
2. Versatile bioadhesives — next-generation surgical adhesives with antimicrobial properties, reversibility, and broad tissue compatibility.
3. Universal Platform for Living Adaptive Toxin-removal (UNI-PLAT) — a programmable synthetic-biology platform for removing toxins from the body.
4. Curative, non-invasive, long-lasting endometriosis therapy — a first-of-its-kind treatment addressing root causes rather than symptoms.
5. ARPA-H Lineage Topic — a dedicated lane for commercializing technologies that prior ARPA-H programs already funded.
6. Rapid comprehensive diagnostic test for multi-system autoimmune disease — fast autoimmune detection usable in primary care.
7. Virtual human brain for neurosurgical robotics — a simulation environment to develop AI-assisted intracranial microsurgery.

Read that list as a strategy document, not a menu. Three of the seven (fertility forecasting, endometriosis therapy, the autoimmune panel) sit squarely in women's health and underserved diagnostic areas that the broader market has chronically underfunded — a signal about where ARPA-H believes private capital has failed. The remaining four are platform and tooling plays. The fit test is binary: if your technology does not map tightly onto one of these seven, this is not your cycle. ARPA-H is not interested in "adjacent" — it wrote these topics because it knows the capability gap it wants closed.

It is a contract, not a grant — and that reframes the whole pitch

Here is the structural fact that trips up grant-trained applicants. ARPA-H issues SBIR awards as contracts, not grants. At NIH, an SBIR grant funds your research and the relationship is relatively hands-off; you pursue the aims and report progress. An ARPA-H contract is a firm-fixed-price instrument with milestone-based, quarterly payments — you get paid as you hit ambitious, pre-negotiated technical milestones, not upfront and not for effort.

That difference cascades into how you should write. A grant reviewer rewards scientific merit and a compelling research question. A contract program manager rewards a credible path to a deliverable on an aggressive timeline. Your Solution Summary needs to convince a program manager that you can hit hard milestones on a clock — that you have de-risked the core technical uncertainty enough to commit to dates. "We will investigate" is grant language. "We will demonstrate X by month six and Y by month twelve" is contract language. ARPA-H explicitly weights a "credible path to market and impact" and the ability to reach a "commercial inflection point," not scientific elegance alone.

The award ceilings reflect the ambition: up to $600,000 for Phase I and up to $3.5 million for Phase II, with Direct-to-Phase II available on the topics that support it (notably the Lineage and autoimmune topics) for teams that can show they have already retired Phase I risk. Direct-to-Phase II is the most efficient route for a company with real preliminary data — it skips the feasibility phase entirely and goes straight for the $3.5M build — but it raises the evidentiary bar at the summary stage. You cannot claim to have skipped feasibility without showing the data that earned the skip.

Eligibility: the rules that quietly disqualify good companies

ARPA-H uses standard SBA small-business rules, and they are where otherwise-strong applicants self-eliminate. You must be a small business with 500 or fewer employees, majority U.S.-owned, performing all work domestically. The work-share thresholds matter and differ by track:

• SBIR: the principal investigator must be more than 50% employed by the small business; the business performs at least two-thirds (66%) of Phase I work and at least half (50%) of Phase II.
• STTR: you must partner with a research institution; the business performs at least 40% and the institution at least 30% of the work, and the PI can be primarily employed by either.

Two operational traps cost teams the deadline every cycle. First, registrations: SAM.gov, the SBA Company Registry, and related certifications can take two weeks or more — if you are not already registered, start today, because an expired or incomplete SAM.gov record will lock you out regardless of how good the science is. Second, foreign affiliation disclosure: ARPA-H runs a strict national-security review, and undisclosed foreign relationships can trigger denial. Disclose proactively and completely; the program treats omissions far more harshly than the relationships themselves.

How to win the Solution Summary

The summary is roughly four to six pages, and its brevity is the trap — short documents are harder, not easier, because there is nowhere to hide a weak thesis. Three things separate the summaries that advance:

Lead with the breakthrough, not the background. ARPA-H explicitly rejects incremental R&D and demands "non-incremental, breakthrough technologies" with "clear technical novelty." The first paragraph should state, in plain terms, what becomes possible with your technology that is impossible today — and why now. If your opening reads like a literature review, you have already lost the program manager's attention.

Quantify the milestones. Because this is a contract, the reviewer is mentally drafting your statement of work as they read. Give them the metrics, the timeline, and the go/no-go gates. Make it easy to imagine paying you in quarterly increments against defined deliverables.

Show the path past the science. ARPA-H pairs awards with an Entrepreneur-in-Residence program and optional I-Corps customer discovery — it cares about reaching a commercial inflection point. Even at the summary stage, name your regulatory pathway, your eventual buyer, and the inflection that the Phase II money buys you. A brilliant technology with no articulated route to patients will lose to a slightly less novel one that can describe how it reaches the clinic.

For the broader federal small-business calendar this summer — including the NSF SBIR/STTR reactivation closing July 27 and the DoD Release 3 window closing July 22 — ARPA-H sits at the high-risk, high-conviction end of the spectrum. It funds fewer, bolder bets and asks you to commit to milestones most agencies would never demand. If your company is building something genuinely transformational in one of those seven areas, the July 10 summary is the single most leveraged page you will write all year. Treat it that way.